Multi-watermarking algorithm for medical image based on KAZE-DCT

• Published in: Journal of ambient intelligence and humanized computing Vol. 15; no. 2; pp. 1735 - 1743
• Main Authors: Zeng, Cheng, Liu, Jing, Li, Jingbing, Cheng, Jieren, Zhou, Jingjun, Nawaz, Saqib Ali, Xiao, Xiliang, Bhatti, Uzair Aslam
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.02.2024; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; Computational Intelligence; Correlation coefficients; Decomposition; Digital imaging; Digital watermarking; Digital watermarks; Engineering; Medical imaging; Neural networks; Original Research; Robotics and Automation; Robustness; Security; User Interfaces and Human Computer Interaction; Wavelet transforms; Medical image; KAZE; Discrete Cosine Transform; Multi-watermarking; Perceptual hashing
• ISSN: 1868-5137; 1868-5145
• DOI: 10.1007/s12652-021-03539-5

With the wide application of digital watermarking technology in the field of medical imaging, the security of medical image information has been improved. Aiming at the problem of poor robustness of medical image watermarking algorithms against geometric attacks, the security of medical images cannot be guaranteed. This paper proposed a zero watermarking algorithm for medical images based on KAZE-DCT. First, KAZE-DCT is used to extract feature vectors of medical images, and perceptual hashing is used to obtain feature sequences of medical images. Then, chaotic mapping is used to encrypt the multi-watermark images, and the zero watermarking technology is applied to embed and extract the watermarks. Finally, the correlation coefficient is used to measure the correlation between the algorithm embedding and extracting the watermarks. Experimental results show that the algorithm can effectively extract watermarks. Moreover, it has good robustness against both common attacks and geometric attacks.